Popular in Professional

Leading Figures

Induction of Social Contagion

Yale University Nicholas A. Christakis

Business

The Allegory of the Favela: The Multifaceted Effects of Socioeconomic Mobility

University of Pennsylvania Leandro S. Pongeluppe

Architecture

Connecting Cities and Nature

University of Virginia Timothy Beatley

Human Rights Law

Human Rights and Corruption: Problems and Potential of Individualising a Systemic Problem

Max Planck Institute Anne Peters

Finance

Chinese Exchange Rate Policy: Lessons for Global Investors

University of California San Diego Michael Melvin

Law

A statistical approach to law school citation rankings

University of Virginia Joshua Fischman

Education

Better Together!

View more insights in Professional - Business - Operational Research

View more figures from Institute for Systems and Computer Engineering, Technology and Science

View the page of Pedro Amorim

Read the Study

Next in Business

Design of Patient Visit Itineraries in Tandem Systems

Nan Liu | Boston College

Mutual and Exclusive: Dyadic Sources of Trust in Interorganizational Exchange

Bill McEvily | University of Toronto

Pedro Amorim highlights how customers typically order multiple things from different sources online.

Image courtesy of interviewee. February 5, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video video is a transcription of a discussion about a study on consumer preferences in e-commerce, particularly regarding delivery methods. The speaker highlights the challenges retailers face in balancing speed of delivery and customer satisfaction. Many retailers prioritize rapid delivery, but the study investigates whether customers actually value consolidated deliveries—receiving multiple items together—even if it takes longer.

The research employed a natural experiment involving an online fashion marketplace that occasionally delivered orders either consolidated or split. Using granular data collected over two years, the study focused on the impact of these delivery methods on the likelihood of product returns. The findings revealed that customers preferred receiving their orders consolidated, resulting in lower return rates, which is significant for retailers facing high costs associated with returns.

This video video preference is explained through psychological mechanisms, particularly the "need for closure," where receiving items together reduces mental tension and frustration associated with staggered deliveries. The study emphasizes the importance for retailers to reevaluate assumptions about customer preferences and to better orchestrate their supply chain to increase consolidated deliveries without incurring excessive costs.

Key managerial implications include the need for data-driven approaches to understand customer preferences and to orchestrate supply chains effectively to enhance delivery methods that align with these insights.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.105 --> 00:03.005
The context from the study actually started a bit

00:03.005 --> 00:04.765
before this, this study.

00:04.945 --> 00:08.685
Um, so I've been interested in this, landscape of,

00:08.685 --> 00:09.805
e-commerce, retail

00:09.905 --> 00:12.405
and, in particular less mile distribution

00:12.405 --> 00:14.085
because as we know, a lot

00:14.085 --> 00:18.885
of retailers are actually struggling to, you know,

00:19.085 --> 00:20.485
square their profit

00:20.545 --> 00:24.405
and loss statements, across different, sectors,

00:24.405 --> 00:27.845
you know, grocery fashion and, and other, and other, and,

00:28.065 --> 00:30.405
and electronics and consumer goods, whatever.

00:30.905 --> 00:33.245
Uh, and one thing that we have, you know,

00:33.245 --> 00:35.365
understood while we were talking with these retailers

00:35.365 --> 00:37.445
and trying to understand how we can help them to actually,

00:37.865 --> 00:41.685
you know, have a better business, we understood

00:41.685 --> 00:45.685
that a lot of assumptions about what customers actually

00:46.275 --> 00:48.365
want were driving their decisions.

00:48.385 --> 00:50.765
Let me give you a very, straightforward example.

00:51.435 --> 00:54.845
Most of these retailers were betting on speed all the time.

00:55.065 --> 00:56.285
We need a fashion retailer.

00:56.385 --> 00:58.245
And they wanted to deliver, you know,

00:58.245 --> 01:01.165
like Farfetch is delivering Gucci bags in 20 minutes.

01:01.165 --> 01:03.525
Farfetch is like the one of the world luxury,

01:03.525 --> 01:05.285
largest luxury e-commerce players.

01:05.625 --> 01:08.725
And then when we look to food retailers like gorillas in

01:08.725 --> 01:11.045
Europe, they were delivering in 10 minutes food.

01:11.465 --> 01:13.405
So a lot of them were banging on speed.

01:13.475 --> 01:15.765
This is of course very expensive.

01:16.345 --> 01:20.205
And our observation is that they didn't really have,

01:20.775 --> 01:23.005
let's say, the, the, we didn't have the knowledge

01:23.005 --> 01:25.765
to understand that, do customers just value speed?

01:25.905 --> 01:28.045
Are there other things that they were valuing?

01:28.505 --> 01:31.605
And while, you know, investigating this question, we came

01:31.665 --> 01:35.845
to understand that, for example,

01:36.425 --> 01:39.285
we were, you know, giving priority

01:39.305 --> 01:42.645
to speed over consolidating goods when you have multiple

01:42.645 --> 01:44.325
packages arriving to your place.

01:44.745 --> 01:47.325
So that is one angle in which we saw, again

01:47.325 --> 01:48.965
and again speed, but I can give you others.

01:49.025 --> 01:51.725
We were, you know, again, putting

01:52.285 --> 01:54.765
priority on speed over precision, for example,

01:54.825 --> 01:57.005
or the precision of the slot in which you deliver.

01:57.625 --> 01:59.925
And we wanted to get a little bit deeper in really

01:59.925 --> 02:02.285
understanding is really speed the sink

02:02.785 --> 02:05.885
or could it be that actually re customers, sorry,

02:05.885 --> 02:07.765
could be actually be interested in something else,

02:08.105 --> 02:09.205
rather than speed.

02:09.385 --> 02:11.685
In this particular study, as I was saying,

02:12.065 --> 02:15.285
we were interested in investigating this difference between,

02:15.465 --> 02:18.365
or, or how, what was the priority that customers were giving

02:18.365 --> 02:21.165
between actually speed and consolidation?

02:21.305 --> 02:23.685
And just to be clear about what consolidation is,

02:23.955 --> 02:26.765
very often when you buy in websites like Amazon

02:27.185 --> 02:30.005
and things like that, what happens is that you actually,

02:30.235 --> 02:31.805
when you order more than one product,

02:32.025 --> 02:35.205
you end up receiving packages in different points in time.

02:35.465 --> 02:37.845
Uh, again, because retailers are prioritized speed.

02:38.385 --> 02:42.325
So what we wanted to do here was actually to have a setting

02:42.965 --> 02:46.445
a retailer that would actually allow us to study this.

02:46.545 --> 02:47.925
Uh, this is tough

02:48.085 --> 02:51.245
because it'll be hard to tell a retailer, well, from now on

02:51.545 --> 02:53.285
for this given amount of customers,

02:53.305 --> 02:56.325
you'll always send it consolidated even if later.

02:56.625 --> 03:00.725
And for another set of of customers you would send as fast

03:00.745 --> 03:02.565
as you can, but not consolidate it.

03:02.625 --> 03:04.765
You know, that's something that would be hard to enforce,

03:04.795 --> 03:08.325
kind of giving a real world experiment of how these things,

03:08.505 --> 03:11.245
would, would be, you know, would we have

03:11.635 --> 03:13.445
what we have to have to actually, you know,

03:13.445 --> 03:15.005
measure the effect that we wanted?

03:15.155 --> 03:18.605
What was, the impact of, consolidating versus speed?

03:19.025 --> 03:22.245
But we were lucky enough to actually find one retailer that

03:23.125 --> 03:26.645
randomly sometimes consolidated orders to customers.

03:26.665 --> 03:29.605
And sometimes they actually didn't consolidate these orders

03:29.745 --> 03:32.965
to customers and they were millions, you know, like

03:33.765 --> 03:37.045
thousands of, of orders that we could learn from

03:37.265 --> 03:40.005
and that we could actually kind of, we already had kind

03:40.005 --> 03:43.005
of a natural experiment that we, we were able

03:43.025 --> 03:45.485
to leverage into, our study.

03:45.825 --> 03:49.245
So basically we used two year granular data, okay?

03:49.245 --> 03:51.725
This was actually an online fashion marketplace

03:52.095 --> 03:56.045
where we were able to really have a causal inference.

03:56.105 --> 03:59.125
So we were able really to understand if consolidating

03:59.865 --> 04:02.525
is actually impacting the customer behavior.

04:02.945 --> 04:05.525
And the way we measure that customer behavior were actually

04:05.635 --> 04:07.485
what was the likelihood of return?

04:07.865 --> 04:09.885
So if we give the products consolidated

04:09.945 --> 04:11.285
or if we give them splits,

04:11.425 --> 04:15.445
but faster, was the customer more likely to return or not?

04:15.465 --> 04:17.205
So that was really the research question

04:17.205 --> 04:22.085
that we tapped into using this, this natural experiment,

04:22.655 --> 04:26.085
which was actually we were lucky enough to have a player

04:26.115 --> 04:30.245
that allowed us to have this quasi random sample of orders

04:30.715 --> 04:34.085
that we were able to, to actually exploit,

04:34.105 --> 04:35.805
in our approach basically.

04:36.105 --> 04:38.805
And so just to give a very visual, idea of

04:38.805 --> 04:41.405
what we did, we were able to match two others,

04:41.495 --> 04:42.885
which were exactly the same.

04:42.995 --> 04:45.165
Like if you have a evil twin

04:45.865 --> 04:48.885
and you were, you know, living on the same place

04:48.945 --> 04:52.525
as your evil twin, but you always got it, you know,

04:52.885 --> 04:56.085
consolidated, but you know, maybe, a little bit later

04:56.225 --> 04:59.525
and your twin was always able to receive it faster,

04:59.985 --> 05:01.205
but split order.

05:01.385 --> 05:03.125
So you are when received the trousers

05:03.125 --> 05:04.205
and the T-shirt together

05:04.545 --> 05:08.165
and your twin was receiving, its the separate,

05:08.505 --> 05:11.165
and we were not only having one twin of Sam,

05:11.165 --> 05:12.685
we were having the twin of Sam, the twin

05:12.685 --> 05:13.885
of Pedro, the twin of Jessica.

05:14.185 --> 05:18.165
So we were having a lot of twins that we were then, able

05:18.225 --> 05:19.845
to measure what was the impact?

05:20.065 --> 05:22.245
Was the even twin returning more

05:22.345 --> 05:24.765
or was the good twin re returning less?

05:24.825 --> 05:28.205
And that was actually what we measured in this,

05:28.205 --> 05:29.765
better together stu study.

05:29.865 --> 05:32.685
As you said, we came without, to be honest,

05:32.685 --> 05:35.325
to the study without not a lot of expectation.

05:35.345 --> 05:38.885
It was not so clear to us, how this would turn out

05:38.885 --> 05:40.965
because again, a lot of people were betting on speed.

05:41.315 --> 05:44.165
Were we really after shaping, I think, you know, so

05:44.165 --> 05:46.005
that was, how we actually started,

05:46.505 --> 05:48.885
but we ended up in a very different situation.

05:48.985 --> 05:52.485
We actually understood that customers prefer

05:52.505 --> 05:53.725
to receive it later,

05:53.905 --> 05:56.965
but consolidated that rather than faster,

05:57.265 --> 05:59.445
but split, we actually understood

05:59.635 --> 06:02.645
that you can lower about one percentage points

06:02.945 --> 06:05.485
of the return rate, which is very important.

06:05.585 --> 06:08.045
As you know, most of these retailers are struggling

06:08.275 --> 06:11.565
with billions of euros of returns skills, right?

06:11.625 --> 06:12.885
So you, this is,

06:12.885 --> 06:16.445
this can actually make a mark not only on the economic side

06:16.445 --> 06:18.725
but also on sustainability side of this business.

06:18.785 --> 06:20.325
But lemme focus again on the results.

06:20.545 --> 06:23.645
We were able to see that actually customers return less.

06:24.035 --> 06:26.325
That actually has an impact on the bottom line

06:26.385 --> 06:27.605
of these retailers

06:27.605 --> 06:30.845
because returning less means that you keep more of the sales

06:30.845 --> 06:32.365
that you actually made.

06:32.825 --> 06:37.445
We saw that an improvement of one point 23 percentage points

06:37.995 --> 06:41.805
when actually if you were to actually send the goods

06:42.445 --> 06:44.325
consolidated instead of split,

06:45.185 --> 06:46.565
it was puzzling for us, right?

06:46.565 --> 06:49.005
Because then, okay, why, why could it be the case?

06:49.005 --> 06:53.045
So entering in the mechanisms, as you said, and, and,

06:53.065 --> 06:54.605
and, and we have actually,

06:54.625 --> 06:56.805
so we are operations management professors.

06:56.835 --> 06:59.645
Basically the ones that were co-author these,

06:59.645 --> 07:02.685
these two papers, what we then went to, okay,

07:02.685 --> 07:04.005
but maybe we need a more more

07:04.005 --> 07:05.605
multidisciplinary approach here.

07:05.605 --> 07:08.885
So we went actually to dig into the psychology literature

07:09.465 --> 07:12.965
and to understand why, why could it,

07:13.155 --> 07:17.365
what could explain some pre, you know, the reversal of, of,

07:17.425 --> 07:19.205
the, the customer's preferences, right?

07:19.205 --> 07:20.325
Because they bought something

07:20.325 --> 07:22.725
and then they are, returning it.

07:22.745 --> 07:24.965
But that depends on the fact that they see it split

07:25.065 --> 07:26.445
or consolidate, okay?

07:26.705 --> 07:29.285
And we saw various different mechanisms

07:29.285 --> 07:30.885
that could actually help to explain.

07:31.385 --> 07:34.205
For example, we know from psychology literature

07:34.205 --> 07:38.085
that if you receive things staggered, they seem longer.

07:38.705 --> 07:41.325
So this could be actually an interesting aspect

07:41.325 --> 07:43.005
because then okay, in the end it's

07:43.005 --> 07:45.125
because you are perceiving it as being, you know,

07:45.305 --> 07:46.325
as taking more time.

07:46.785 --> 07:49.125
So there were a lot of mechanisms that we understood.

07:49.185 --> 07:51.525
And the great thing is that with our data, we were able

07:51.525 --> 07:53.045
to test this mechanism, okay?

07:53.045 --> 07:55.205
We were able to see if, for example,

07:55.385 --> 07:59.245
and focusing on this case of multiple orders, we could see

07:59.245 --> 08:01.845
that, you know, then if this mechanism was in place,

08:02.225 --> 08:06.325
having four splits would be worse than having two splits.

08:06.465 --> 08:09.605
And we didn't really get into that, conclusion.

08:10.235 --> 08:13.845
What we don't find was that found, sorry, was

08:13.845 --> 08:15.765
that there was another mechanism

08:15.795 --> 08:17.085
that could really explain that.

08:17.185 --> 08:18.925
And that mechanism is

08:18.925 --> 08:21.525
what we commonly call the need for closure.

08:22.185 --> 08:25.445
Let me give you an example. You have your to-do list, right?

08:25.785 --> 08:28.005
And we all feel some pressure while we are having

08:28.005 --> 08:29.245
these two go lists, right?

08:29.415 --> 08:32.365
While we don't really close some of the deals

08:32.365 --> 08:35.885
that we have there, there's always a mental, mental tension

08:35.995 --> 08:39.645
that we keep going on, with us with along the time.

08:40.105 --> 08:42.845
So that is actually on, the literature,

08:42.985 --> 08:44.205
the psychology literature.

08:44.265 --> 08:46.125
We call that the zagar effect.

08:46.225 --> 08:49.325
So that's an effect of having dismantle folder while you

08:49.325 --> 08:51.925
still have an active test that you didn't really conclude.

08:52.455 --> 08:54.525
Let's get back to our retail setting.

08:54.825 --> 08:58.045
You know, you ordered something at one time, your trousers

08:58.045 --> 09:00.445
and a blouse or you know, a shirt or whatever.

09:01.625 --> 09:03.885
One thing is that when you get it consolidated, you know,

09:03.985 --> 09:06.765
you, you scratch out of your to-do list right away.

09:07.235 --> 09:10.205
Whereas if you get it, your trousers today

09:10.305 --> 09:12.685
and your sweater in three days, you see,

09:13.105 --> 09:15.845
you still carry some mental tension along the way.

09:16.545 --> 09:18.325
And guess what? It turns out that

09:18.325 --> 09:21.045
that mental tension leads into some frustration

09:21.065 --> 09:22.205
and you just,

09:22.345 --> 09:25.885
return more than you will return if you get it,

09:26.325 --> 09:28.725
altogether consolidated, as I said in the beginning.

09:28.825 --> 09:31.205
So that was really the mechanism that we found

09:31.205 --> 09:33.085
that was driving the results.

09:33.385 --> 09:34.685
And that was very interesting

09:34.685 --> 09:37.365
because we were even to get one step, you know,

09:37.545 --> 09:39.325
in deeper and understanding

09:39.325 --> 09:42.445
that the more the goods are apart, so

09:43.525 --> 09:45.305
two situations you receive on day one

09:45.305 --> 09:47.665
and you receive on day three, or you receive in day one

09:47.725 --> 09:51.785
and on day five, if you do receive it one day, one

09:51.785 --> 09:53.545
and day five, you are more likely

09:53.685 --> 09:56.985
to return than if you would receive, you know, still split

09:57.285 --> 09:58.345
but closer together

09:58.345 --> 10:01.145
because you carry that tension more along the way.

10:01.325 --> 10:03.665
So I think the first managerial implication is that

10:04.725 --> 10:07.825
we are urging retailers to stop assuming things about

10:08.205 --> 10:11.865
how customers actually, what do you actually want?

10:12.085 --> 10:14.945
And you better, you know, find ways to actually measure

10:15.015 --> 10:16.985
that either like we did in our,

10:17.105 --> 10:18.785
I would call it more data-driven approach,

10:19.165 --> 10:22.345
but maybe even asking to be something as it starts to see

10:22.345 --> 10:23.705
where they, they actually are.

10:23.975 --> 10:27.065
Because what happens is that retailers are committing a lot

10:27.065 --> 10:29.705
of resources, spending a lot of money in things

10:29.775 --> 10:32.625
that actually may turn out not to be what customers want.

10:32.625 --> 10:35.625
So I would say that is a very broad first managerial insight

10:35.625 --> 10:37.865
that, you know, starts broadly important, right?

10:37.865 --> 10:41.385
Let's start that way, but going more specific into our,

10:41.415 --> 10:44.545
into our results, I think what we are really calling out is

10:44.545 --> 10:48.265
that retailers, they need to orchestrate their supply chain.

10:48.575 --> 10:51.665
It's not enough to really, you know, get an order

10:51.885 --> 10:54.025
and, you know, sense to the disease

10:54.085 --> 10:56.265
or to the suppliers that are feeling that orders,

10:56.725 --> 10:58.865
but there's a need of some orchestration

10:59.125 --> 11:01.465
and that orchestration needs to be such

11:01.815 --> 11:04.225
that you are increasing the likelihood of getting

11:05.585 --> 11:07.425
consolidated shipping to your customer.

11:07.775 --> 11:09.705
Okay? Where you are doing that consolidation,

11:09.985 --> 11:13.225
I think it'll depend a lot on the network structure

11:13.405 --> 11:15.225
of the retailer that we are talking with.

11:15.285 --> 11:17.825
So in our case, we were partnering with a marketplace,

11:17.925 --> 11:20.465
so basically they were connecting suppliers

11:20.465 --> 11:22.305
and customers with direct shipment.

11:22.815 --> 11:26.985
That means that maybe instead of sending right away the,

11:27.125 --> 11:29.905
the, the orders to be fulfilled with different suppliers,

11:30.315 --> 11:33.905
maybe they need to wait a bit in one hand or leave the,

11:33.965 --> 11:37.865
or, you know, or expedite the other to make sure that

11:38.485 --> 11:42.065
the likelihood of having a consolidated delivery at the end

11:42.155 --> 11:43.265
would, would increase.

11:44.205 --> 11:48.505
Of course, you can also enforce that, for example, talking

11:48.525 --> 11:50.465
to your third party logistics provider

11:50.485 --> 11:52.785
and say, Hey, everything that

11:52.785 --> 11:55.565
Goes to this customer, you hold it in your hub,

11:55.665 --> 11:57.605
you do whatever you want, but you consolidate.

11:57.625 --> 12:01.125
And by the way, some three pls are already doing that, like,

12:01.125 --> 12:03.405
for example, DHL of offer these services.

12:03.985 --> 12:05.365
But of course there's the cost to that.

12:05.425 --> 12:07.485
So I guess there's also the need

12:07.545 --> 12:10.005
and not to rush into, you know, doing things

12:10.005 --> 12:13.125
that then they don't pay off to balance the fact

12:13.125 --> 12:15.525
that you have actually low re lower returns

12:15.745 --> 12:17.725
to the adding costs that some

12:17.725 --> 12:20.405
of the solutions may actually have.

12:20.785 --> 12:23.645
So being it, so in a nutshell, you can, I think you,

12:23.825 --> 12:26.165
we are saying that you can do it in a more soft approach,

12:26.845 --> 12:28.485
orchestrating more the supply chain

12:28.505 --> 12:30.005
to increase the likelihood

12:30.085 --> 12:31.485
of getting a consolidated delivery

12:31.905 --> 12:34.725
or you can enforce it, orchestrating comes for free.

12:34.915 --> 12:37.605
It's more like putting some sort into your process

12:37.705 --> 12:39.125
and make sure that that happens.

12:39.625 --> 12:41.885
If you are going for like a full enforcement,

12:42.165 --> 12:44.885
I guess there's some additional calculations

12:44.885 --> 12:45.845
that you actually need to do.

Examining E-Commerce Delivery Preferences

00:00:00 - The context from the study actually started a bit before this, this study. so I've been interested in this, landscape of, e-commerce, retail and, in particular less mile distribution because as we know, a lot of retailers are actually struggling to, you know, square their profit and loss statements, across different, sectors, you know, grocery fashion and, and other, and other, and, and electronics and consumer goods, whatever.
00:00:33 - understood while we were talking with these retailers and trying to understand how we can help them to actually, you know, have a better business, we understood that a lot of assumptions about what customers actually want were driving their decisions. Let me give you a very, straightforward example. Most of these retailers were betting on speed all the time. We need a fashion retailer. And they wanted to deliver, you know, like Farfetch is delivering Gucci bags in 20 minutes.
00:01:03 - largest luxury e-commerce players. And then when we look to food retailers like gorillas in Europe, they were delivering in 10 minutes food. So a lot of them were banging on speed. This is of course very expensive. And our observation is that they didn't really have, let's say, the, the, we didn't have the knowledge to understand that, do customers just value speed? Are there other things that they were valuing? And while, you know, investigating this question, we came
00:01:36 - we were, you know, giving priority to speed over consolidating goods when you have multiple packages arriving to your place. So that is one angle in which we saw, again and again speed, but I can give you others. We were, you know, again, putting priority on speed over precision, for example, or the precision of the slot in which you deliver. And we wanted to get a little bit deeper in really understanding is really speed the sink
00:02:05 - could be actually be interested in something else, rather than speed. In this particular study, as I was saying, we were interested in investigating this difference between, or, or how, what was the priority that customers were giving between actually speed and consolidation? And just to be clear about what consolidation is, very often when you buy in websites like Amazon and things like that, what happens is that you actually, when you order more than one product, you end up receiving packages in different points in time.
00:02:38 - So what we wanted to do here was actually to have a setting a retailer that would actually allow us to study this. this is tough because it'll be hard to tell a retailer, well, from now on for this given amount of customers, you'll always send it consolidated even if later. And for another set of of customers you would send as fast as you can, but not consolidate it. You know, that's something that would be hard to enforce, kind of giving a real world experiment of how these things,

Impact of Consolidation Versus Speed in Retail

00:03:11 - what we have to have to actually, you know, measure the effect that we wanted? What was, the impact of, consolidating versus speed? But we were lucky enough to actually find one retailer that randomly sometimes consolidated orders to customers. And sometimes they actually didn't consolidate these orders to customers and they were millions, you know, like thousands of, of orders that we could learn from and that we could actually kind of, we already had kind
00:03:43 - to leverage into, our study. So basically we used two year granular data, okay? This was actually an online fashion marketplace where we were able to really have a causal inference. So we were able really to understand if consolidating is actually impacting the customer behavior. And the way we measure that customer behavior were actually what was the likelihood of return? So if we give the products consolidated or if we give them splits,
00:04:15 - So that was really the research question that we tapped into using this, this natural experiment, which was actually we were lucky enough to have a player that allowed us to have this quasi random sample of orders that we were able to, to actually exploit, in our approach basically. And so just to give a very visual, idea of what we did, we were able to match two others,
00:04:42 - Like if you have a evil twin and you were, you know, living on the same place as your evil twin, but you always got it, you know, consolidated, but you know, maybe, a little bit later and your twin was always able to receive it faster, but split order. So you are when received the trousers and the T-shirt together and your twin was receiving, its the separate, and we were not only having one twin of Sam, we were having the twin of Sam, the twin
00:05:14 - So we were having a lot of twins that we were then, able to measure what was the impact? Was the even twin returning more or was the good twin re returning less? And that was actually what we measured in this, better together stu study. As you said, we came without, to be honest, to the study without not a lot of expectation. It was not so clear to us, how this would turn out because again, a lot of people were betting on speed. Were we really after shaping, I think, you know, so
00:05:46 - but we ended up in a very different situation. We actually understood that customers prefer to receive it later, but consolidated that rather than faster, but split, we actually understood that you can lower about one percentage points of the return rate, which is very important. As you know, most of these retailers are struggling with billions of euros of returns skills, right? So you, this is, this can actually make a mark not only on the economic side

Impact of Order Consolidation on Returns

00:06:18 - But lemme focus again on the results. We were able to see that actually customers return less. That actually has an impact on the bottom line of these retailers because returning less means that you keep more of the sales that you actually made. We saw that an improvement of one point 23 percentage points when actually if you were to actually send the goods consolidated instead of split, it was puzzling for us, right?
00:06:49 - So entering in the mechanisms, as you said, and, and, and, and we have actually, so we are operations management professors. Basically the ones that were co-author these, these two papers, what we then went to, okay, but maybe we need a more more multidisciplinary approach here. So we went actually to dig into the psychology literature and to understand why, why could it, what could explain some pre, you know, the reversal of, of,
00:07:19 - Because they bought something and then they are, returning it. But that depends on the fact that they see it split or consolidate, okay? And we saw various different mechanisms that could actually help to explain. For example, we know from psychology literature that if you receive things staggered, they seem longer. So this could be actually an interesting aspect because then okay, in the end it's because you are perceiving it as being, you know, as taking more time. So there were a lot of mechanisms that we understood.
00:07:51 - to test this mechanism, okay? We were able to see if, for example, and focusing on this case of multiple orders, we could see that, you know, then if this mechanism was in place, having four splits would be worse than having two splits. And we didn't really get into that, conclusion. What we don't find was that found, sorry, was that there was another mechanism that could really explain that. And that mechanism is what we commonly call the need for closure.
00:08:25 - And we all feel some pressure while we are having these two go lists, right? While we don't really close some of the deals that we have there, there's always a mental, mental tension that we keep going on, with us with along the time. So that is actually on, the literature, the psychology literature. We call that the zagar effect. So that's an effect of having dismantle folder while you still have an active test that you didn't really conclude.
00:08:54 - You know, you ordered something at one time, your trousers and a blouse or you know, a shirt or whatever. One thing is that when you get it consolidated, you know, you, you scratch out of your to-do list right away. Whereas if you get it, your trousers today and your sweater in three days, you see, you still carry some mental tension along the way. And guess what? It turns out that that mental tension leads into some frustration and you just, return more than you will return if you get it,

Optimizing Delivery: Consolidation vs. Speed

00:09:28 - So that was really the mechanism that we found that was driving the results. And that was very interesting because we were even to get one step, you know, in deeper and understanding that the more the goods are apart, so two situations you receive on day one and you receive on day three, or you receive in day one and on day five, if you do receive it one day, one and day five, you are more likely to return than if you would receive, you know, still split
00:09:58 - because you carry that tension more along the way. So I think the first managerial implication is that we are urging retailers to stop assuming things about how customers actually, what do you actually want? And you better, you know, find ways to actually measure that either like we did in our, I would call it more data-driven approach, but maybe even asking to be something as it starts to see where they, they actually are. Because what happens is that retailers are committing a lot of resources, spending a lot of money in things
00:10:32 - So I would say that is a very broad first managerial insight that, you know, starts broadly important, right? Let's start that way, but going more specific into our, into our results, I think what we are really calling out is that retailers, they need to orchestrate their supply chain. It's not enough to really, you know, get an order and, you know, sense to the disease or to the suppliers that are feeling that orders, but there's a need of some orchestration and that orchestration needs to be such
00:11:05 - consolidated shipping to your customer. Okay? Where you are doing that consolidation, I think it'll depend a lot on the network structure of the retailer that we are talking with. So in our case, we were partnering with a marketplace, so basically they were connecting suppliers and customers with direct shipment. That means that maybe instead of sending right away the, the, the orders to be fulfilled with different suppliers, maybe they need to wait a bit in one hand or leave the,
00:11:38 - the likelihood of having a consolidated delivery at the end would, would increase. Of course, you can also enforce that, for example, talking to your third party logistics provider and say, Hey, everything that Goes to this customer, you hold it in your hub, you do whatever you want, but you consolidate. And by the way, some three pls are already doing that, like, for example, DHL of offer these services.
00:12:05 - So I guess there's also the need and not to rush into, you know, doing things that then they don't pay off to balance the fact that you have actually low re lower returns to the adding costs that some of the solutions may actually have. So being it, so in a nutshell, you can, I think you, we are saying that you can do it in a more soft approach, orchestrating more the supply chain to increase the likelihood of getting a consolidated delivery or you can enforce it, orchestrating comes for free.

Ensuring Effective Enforcement in Operations

00:12:44 - and make sure that that happens. If you are going for like a full enforcement, I guess there's some additional calculations that you actually need to do.

Faculti AI can make mistakes

Research

Policy

Analysis